Lasers measure wind velocity at long distance to improve mathematical models

Author: EIS Release Date: Aug 31, 2023


Danish company Vind-Vind is developing a turbulence model for the effect of wind on structures in complex environments.
 
Spectrum 3d wind simulation
It will eventually include the atmosphere high above the volume of interest with wind gusts from different directions.
 
As such, said Vind-Vind CEO Per Jørgensen, wind-tunnel measurements are not good enough and he needed real-world measurements throughout a large volume of atmosphere to improve and verify the model.
 
 
“At present, there are two ways to measure wind movement, either low resolution at several kilometres, or high resolution over a few hundred meters,” he said, so “we created a new lidar instrument to measure long distances at a high resolution”.
 
 
The result was a laser scanner that projectes 10ns pulses into the volume of interest, with reflections from dust particles processed to extract their return Doppler shift to extract velocity.
 
Do do the necessary calculations, the company planned to use an FPGA platform but rejected it as being too complex to programme and not having enough compute power to handle the large amount of data being created every second, according to German digitiser company Spectrum, which brought this tail to Electronics Weekly’s attention because it was to Spectrum that Vind-Vind turned to replace its FPGA-based number crunching idea.
 
Spectrum Digitizer-M5i.3321
 
Instead, Vind-Vind used a 12bit 3.2Gsample/s Spectrum M5i.3321-x16 digitiser (left, essentially an ADC with a 16 lane PCIe interface) in a PC along with a 6,144 core Nvidia Cuda Quadro A4000 GPU card to do the maths, with Spectrum’s ‘SCAPP'(Spectrum Cuda access for parallel processing) software interfacing between the ADC and processing cards at up to 12.8Gbyte/s.
 
The data rate was more than needed, “but it gives us the margin to allow for noisy conditions and weak signals”, said Jørgensen. “The extra bandwidth also means that we can immediately identify and filter out high frequency noise leaving only low frequency noise to be eliminated later when processing the data.”
 
Vind-Vind envisages the final version of its model being used in situations where atmospheric turbulence interacts with urban environments, clusters of wind turbines, bridges or airports. A sister company, PJ Science, will produce and sell the lidar system.